GSM GPRS Module at 1.93$!!!!!

Surfing on ebay I found this:

http://www.ebay.com/itm/M590E-GSM-GPRS- … SwYlJW6pub

Ebay sellers say it is 5V powered, but after some googling seems that the right power is from 3.6 V to 4.3 V and seems compatible with 3,3V MCUs.

Here some reference:

http://www.sunduino.pl/wordpress/gsm-m5 … -i-prosto/
http://www.radioterminal.ru/upl_instruc … s_V1.0.pdf
http://wless.ru/files/GSM/Neoway/Neoway … e_V1_0.pdf
http://www.academia.edu/15633281/GSM_GP … oway_M590E

I’ve just ordered one (not possible to resist… )

So if anyone is interested…

Cheers

Thanks for info, ordered 2

Pito wrote:Thanks for info, ordered 2

RogerClark wrote:Pito wrote:Thanks for info, ordered 2

BTW You can probably save a few of cents by purchasing this one. Price: US $1.82

I’ve been tempted too … But reluncant to have to subscribe for a SIM card… (I still don’t own any mobile, although I had one in 2000)

…………

GSM/GPRS. In the US, that’s mainly AT&T. Verizon has never used GSM.
I suspect most of the EU is ending it sooner.

http://www.sine-wave.com/blog/2g-sunset … y5Z_nprWYg

stevech wrote:GSM/GPRS. In the US, that’s mainly AT&T. Verizon has never used GSM.
I suspect most of the EU is ending it sooner.

http://www.sine-wave.com/blog/2g-sunset … y5Z_nprWYg

ahull wrote:stevech wrote:GSM/GPRS. In the US, that’s mainly AT&T. Verizon has never used GSM.
I suspect most of the EU is ending it sooner.

http://www.sine-wave.com/blog/2g-sunset … y5Z_nprWYg

stevech wrote:
I suspect most of the EU is ending it sooner.

even at 8c per transmission, you would need to be careful what your IoT device sends.

I thought, IoT, generally refers to Internet connected devices, rather than something that sends texts etc via GSM

…………

despite many requests and removal warnings, the bane of Snooker World Championship 2016 was mobiles ringing.
‘oh this session is all safety play and will run rather late’, i’d better delay the heating.
so if anything was sent back to the mobile in number of rings code, you’d be ejected moderately slowly

stephen

Please do not steal data service… I’ve seen people go to jail…

Ray

Ray

I agree.

Posts have been deleted.

Ray can you remove your quote, it propagates the information

………..

Simon

This is somewhat off topic, even for the this thread.

I’d rather it was not discussed here at all.

I’m not suggesting you stop posting completely, but we all have to self moderate.

Thanks.

RogerClark wrote:Simon

This is somewhat off topic, even for the this thread.

I’d rather it was not discussed here at all.

I’m not suggesting you stop posting completely, but we all have to self moderate.

In the telephone crossbar switch central office days, someone found that if, on an incoming call, a DIY device goes off-hook for 1/4 second, the circuit to the caller is made but the toll gear (billing system) didn’t register the call. (From the days of toll calls for most any call farther than 50 miles or so.)

@Steve:

… Maybe a different case than the one you are thinking, but the IT manager at the South Carolina textile plant that implemented the technique (circa 1978) lost his job, became a convicted felon, and screwed up the rest of his life for doing this stupid, stupid thing! What is so sad, the money he tried to save was not even his own. Fraud is bad business for everyone.

Ray
at&t Senior IT Enterprise Architect, retired

mrburnette wrote:…at&t Senior IT Enterprise Architect, retired

I have such a module. Works very good. It has an internal diode to reduce VCC from 5 to 4.3V. A prepaid simcard doesent cost a lot. $20 here, including $20 prepaid balance. AT commands are in “Neoway M590 AT Command Sets_V3.0.pdf”

This device WONT BOOT unless you connect the BOOT pin. I did just route it to DTR and this works.

This device can do SMS send / receive. It can also do tcp-ip. I have no data plan and i have no idea what it costs me, but i can connect to my PC via GPRS and exchange data with a $2 phone worldwide. This did pretty much blow my mind. I did not test if DNS does work, i did connect to fixed IP, but it should work.

You need to know the IP stuff and the NAT stuff on your router to get it working. The sequence for TCP-IP is:
at+xisp=0 // interner tcpip stack verwenden
at+cgdcont=1,"ip","internet" // sunrise apn setzen
at+xiic=1 // start ppp
at+tcpsetup=0,177.194.29.176,1005 // call meine ip, port 1005
at+tcpsend=0,10 // sende 10 zeichen
// send 10 char here


This module would require a separate topic for networking beginners like me
Like:
1. how to wire the module (where to connect boot – we do not have dtr)
2. how to power the module
3. how to write a sketch to send a sms
4. how to write a sketch to receive a sms
5. etc..

Pito wrote:1. how to wire the module (where to connect boot – we do not have dtr)

So the power source should give 1A I guess.
DTR – there is no dtr on the 1.93$ module. Most probably we will run it from maple mini or such..
What is ssWrite?

Pito wrote:So the power source should give 1A I guess

The .pl link in the first post recommends:
1. to use power source with at least 2A current capability
2. Boot pin to Gnd
3. To replace the 4k7 resistor with a 10k (sim card signal pullup)..
The .edu source says the powering is critical – a 3.6V LiPo and 1000uF decoupling, powering on/off the module and the power source sequence is critical.
Also the voltage shall not drop below 3.3V while txing.
That would not work reliably with LiPo 3.6V and the 1N4001 diode there however, as the diode will have about 0.7V forward drop at 2A (1n4001 is 1A diode, btw).. Maybe schottky may help, or none..

Yep, i guess the diode is a cheapo hack to run it on 5V. Must be removed operating on 1 cell lipo.
I run it on a 2A 12V source with cheapo chinese step-down to 4.5V, prob 4A max. No problems.

I did sucesfully send an SMS
Setup, text mode, gsm charset
at+cmgf=1
at+cscs="GSM"


This, for low quantity, might be viable. Maybe not.

http://neo.aeris.com/
http://neo.aeris.com/pricing/plans-pricing/

Aeris has been doing low rate telemetry for a very long time.

RogerClark wrote:Pito wrote:Thanks for info, ordered 2

mine have arrived.
looking at the board, diode (bar) track indicate which end of the big cap is positive.
someone remind me, is the bar line on the capacitor is positive or negative?
i have this funny feeling its the negative.

stephen

zmemw16 wrote:mine have arrived.
looking at the board, diode (bar) track indicate which end of the big cap is positive.
someone remind me, is the bar line on the capacitor is positive or negative?
i have this funny feeling its the negative.

stephen

oh good, no bang then.
it would not have been funny, although i’m tempted to sacrifice one to see

srp

zmemw16 wrote:oh good, no bang then.
it would not have been funny, although i’m tempted to sacrifice one to see

srp

There is a signe on the underside of the LED indicating polarity.

i hadn’t thought about led polarity

i suspect the module is 2mm spacing and a whole different technique.

pureed water melon – um!
6uF – excellent

i had to open the video flash frame in another window for it to play, no entry sign in the post.

stephen

Those modules are cheaper at aliexpress

They are listing at $1.55 with free P&P, almost 20% discount over $1.93

I struggle to understand how they can pack and post anything for that.

I’m resisting until I’ve seen a bit more feedback here.

Also, I like the look of the SIM800L quad-band modules.
Insanely expensive at $4.40, but Quad-band for goodness sake

gbulmer wrote:Those modules are cheaper at aliexpress

Also, I like the look of the SIM800L quad-band modules.
Insanely expensive at $4.40, but Quad-band for goodness sake

simonf wrote:
If you are thinking of using this in the UK you only need 900/1800Mhz, but beware whose SIM you use you cant use “Three” sims as they wont work even though they can fall back to 2G. There are some other networks that use the “three” network which should also be avoided.

See http://kenstechtips.com/index.php/mobil … d-coverage

gbulmer wrote:

We do try to make stuff that works in most countries, not just the UK.

Thank you very much again for your helpful links and cost analysis.

The target users are schools, colleges and universities, so avoiding cost shocks is important. Hence my interest in receive only, no-top-up SIMs.

gbulmer wrote:Thank you very much again for your helpful links and cost analysis.

The target users are schools, colleges and universities, so avoiding cost shocks is important. Hence my interest in receive only, no-top-up SIMs.

Thank you simonf. I think that is a solution worth trying; maybe not a 100 SIMs at first

How about robot control over SMS using the same SIMs?-) Network latency might be a problem!

Do you live in the UK? What part? I owe you a pint.
Cheers.

gbulmer wrote:Thank you simonf. I think that is a solution worth trying; maybe not a 100 SIMs at first

I received two modules and it arrived last week. I didn’t bother to look at it closely, gave the seller good feedback and put the package into chinese junk box.

Now, I took better look at it. TLDR – OH MY, IT IS CRUSTY AND UGLY PIECE OF CRAP!
The modules are apparently desoldered from some GSM equipment – though listing says “New: A brand-new, unused, unopened, undamaged item in its original packaging (where packaging is applicable).”
The bottom side of module is bulged
https://picasaweb.google.com/lh/photo/4 … directlink
indication serious overheating of the PCB and its delamination. There are flux residuals
https://picasaweb.google.com/lh/photo/I … directlink
Upper side of PCB is crusty as hell, full of dust
https://picasaweb.google.com/lh/photo/W … directlink
and as you can see, the passive network around GSM RF output is damaged.
The sticker on upper side of module looks like removed or cooked during desoldering process. Bleh
https://picasaweb.google.com/lh/photo/R … directlink

second piece is somehow better, but not great at all
https://picasaweb.google.com/lh/photo/c … directlink
desoldered as well
https://picasaweb.google.com/lh/photo/c … directlink

The modules are of different revisions – 1.0 vs 1.3, not sure what differences are there. I’m pretty sure the crusty module will not work, the another one – perhaps.
I noticed the 1N4007 diodes have leads approximately 0,5mm thick. Cost cutting, here we go. The power supply is 1N4007 in series with module, decoupled with single 100uF capacitor – now this is so wrong…

Regarding the general hints for GSM modules:
– good power supply. Usually the modules expect one Li-Ion battery, so 3.6-4.2V. Voltages over 4.5V are often indicated as overvoltage and module can shut down. The same goes for undervoltage. The 1N4007 in series with 5V is joke.
– good power supply, again. Use proper decoupling, even with beefy source as Li-Ion battery. Few pieces of 100uF tantalum capacitors and one 10uF ceramics can’t hurt.
– beware of RF interference. buy a lot of RF chokes, 100pF capacitors, use at least doublesided PCB and good grounding
– Sending one SMS is simple, making good AT commands parser is nightmare.
– the modules usually do have 2,8V or 3,3V signalling voltages.

I’ve done a few GSM phones before
https://picasaweb.google.com/1118907411 … PhoneTest1
https://hackaday.io/project/558-rotary- … bile-phone
https://picasaweb.google.com/1001130002 … tK_Wi6D-fw
and except of the rotary dial phone it was one of the more complicated side of my projects.

jaromir wrote:I received two modules and it arrived last week. I didn’t bother to look at it closely, gave the seller good feedback and put the package into chinese junk box.

I have build 2 of those modules. The parts were NEW, not used. You must have had bad luck with your supplier.

I run both boards as delivered, no changes at all. I use the included antenna (that crappy coiled thing). I can send / receive SMS and i can get TCP-IP connections over GPRS. Surprisingly the module has BETTER reception than my smartphone (same provider, weak gsm signal).

Dont scare people off. This sub $2 things work excellently well. If you dont need audio there is no reason to get the $6 sim800l or the $20 sim900 modules.

I’m not scaring people off, I just described what I received. If it scares people off, then it is not my mistake.

Cesco wrote:I have build 2 of those modules. The parts were NEW, not used. You must have had bad luck with your supplier.

I run both boards as delivered, no changes at all. I use the included antenna (that crappy coiled thing). I can send / receive SMS and i can get TCP-IP connections over GPRS. Surprisingly the module has BETTER reception than my smartphone (same provider, weak gsm signal).

Dont scare people off. This sub $2 things work excellently well. If you dont need audio there is no reason to get the $6 sim800l or the $20 sim900 modules.

ahull: brand new modules should have goldplated pins, with no solder on it. Is it your case?
Mine are definitely desoldered and I’m pretty sure it’s not customer return – it is obvious the modules have been sitting on dusty and moist PCB for months or years and then desoldered.
I’ll give it one more try – it seems like modules from DX are marginally cheaper and I have better experience with DX quality. I just had bad luck with this seller.

jaromir wrote:ahull: brand new modules should have goldplated pins, with no solder on it. Is it your case?

A have ordered two and one was used (desoldered) as well. And why I don’t care? It was cheap and it works, so I it fine. Few years ago I have been used mostly desoldered parts so it is like return to my childhood ))

The AI Tinker A6 dev board seems like a better option @ US$4.2 if you want consistency (Although I have not had chance to try one yet). As the AI Thinker A6 is a new (Recently Released) device I would not of thought there could be second hand parts on the market yet. For US$ 4.2 you also get a “proper” aerial although it could be just a piece of plastic with nothing in it.

The price has been falling over the past weeks already, maybe AI Thinker are looking to corner the Wide Area Wireless hobby market like they have done with the ESP8266 in the WIFI Market. To introduce a device > $4 that is quad band and audio is no mean feat.

Any links to the AI Thinker A6 module documentation, and / or pinout of the breakout board?

Kenjutsu wrote:Any links to the AI Tinkerer A6 module documentation, and / or pinout of the breakout board?

In roughly the time it took to heat my pizza… behold…

No prizes for the soldering I must admit , and since I couldn’t find my magnifier, or my leaded solder, or indeed my tweezer set (they are all presumably hiding in the same place), I had to use my leadfree crappy solder, two pairs of reading glasses, and a bigger tipped iron to be sure of melting the crappy solder. I’ll test it in the morning. The other two will be neater, I promise, once I find the missing tools.

EDIT: I presume I can link out that diode and run it straight from a LiPo… time will tell.. I can see at least one of these joining the magic smoke club.

Despite the silkscreen on the board bragging of 5V TTL UART, I also suspect that the board might not be too keen on being driven directly by 5V serial data. The Neoway reference manual suggests that the part is designed for use with LiPos and that input voltage in excess of 4.8V may be too much for the thing to bear. http://cyntech.co.uk/downloads/neoway-m … ual-v1.pdf

Since I intend to run it from a LiPo, I don’t think this is likely to be an issue for me, but it is something to be taken in to account when connecting the thing up. I would assume 3V3 logic rather than 5V is likely to be your best bet.

Also …“VBAT (Pin2 and 3) is the main power supply of the module, with input range of 3.3V to 4.8V, with the recommended voltage of 3.9V. Maximum instantaneous current input to the module is 2A. USE HIGH VALUE, LOW ESR CAPACITORS, CLOSE TO THE MODULE. “… in other words, you can’t drive the module from a tiny little LDO voltage regulator, ‘cos it will vanish in a puff of smoke. If you use a voltage regulator it must be beefy enough to cope with 2A without complaining. Furthermore if you intend to switch the module on and off with your main MCU, to conserve battery life, you will need to use a mosfet or transistor that can switch 2A plus.

ahull wrote:Despite the silkscreen on the board bragging of 5V TTL UART,

> According to the spec sheet it uses 2.85V logic Levels but can tolerate upto 3.3v.

I am using mine with 3.3V logic and have no problems.
I dont know if the device tolerates 5V logic from the arduino-avr devices. Has someone tested this?

I modules arrived the other day, but in closer inspection, the GPRS module has been reclaimed and is not New (so is not as described in the eBay listing)

So I gave negative feedback to the eBay vendor as I can’t be bother arguing with them, and the changes are they are a fly-by-night vendor and don’t care what feedback is given.

I have not had time to solder the thing together and I will need to clean the solder blobs off the GPRS module first, otherwise its going to be hard to solder it down neatly, as there are blobs of solder on some pads but not all of them

RogerClark wrote:I modules arrived the other day, but in closer inspection, the GPRS module has been reclaimed and is not New (so is not as described in the eBay listing)

So I gave negative feedback to the eBay vendor as I can’t be bother arguing with them, and the changes are they are a fly-by-night vendor and don’t care what feedback is given.

If you put “M590E” into g…e you get hundreds of pictures of that module, with various motherboards, etc. All modules are de-soldered from a scrap equipment, sure.
@ahull – maybe a little bit better antenna would help. The one there is rather a “dummy load”. There are antenna modules – $1 ceramic patch antennas, which may work even in Scottish Highlands (I traveled across 20y back..).

Since the datasheet specifically mentions “AMR with data-only” and since I assume AMR in this instance refers to automated or automatic meter reading I suspect the devices are recovered from some sort of very common power, water or electricity meter device.

This would explain the large number of recovered units in the e-waste stream. This in turn suggests that the majority, if not all of these will, like the large number of “Arduino Nokia 5110 display” ebay items, come from salvage, rather than surplus stock. I have no problem with that, but may be of concern to others.

The external antenna looks to be a must. I’ll see if I can solder on an RF pigtail, and design some sort of yagi antenna. I know where the nearest cell tower is, so I have a good line of sight to it, its just rather too far for the bent paper clip that is supplied with the board.

If you want to play with antenna design (and this may not be entirely legal in your area, so proceed with caution), you can take a look at

http://www.k7mem.com/Electronic_Noteboo … i_vhf.html

Example. “The current design is an antenna for 1800.000 MHz. It has 1 Reflector, 1 Driven, and 14 Director Elements. Estimated Gain is 14.234 dBd”. … that will probably allow me to connect to the moon. the advantage of the relatively high frequency of course is that a 14 element yagi antenna is only about two feet long, and not twenty feet.

Despite what I wrote about the modules I received, I gave it a try. Soldered both of them.
The first one, with crusty and overheated module is dead as brick.
The second one, works, somehow. After “booting” (holding BOOT pin for a second on low level) the BOOT LED stars blinking and module accepts AT commands at 9600 baud.
All I wrote before, about need for stable and beefy power supply is still valid. The original way of generating power for module by dropping 0,7V on diode is joke (module restarted after entering PIN – due to current spikes for network registration), so I connected stable 4,0V power directly across the filtering capacitor. By the way, the diode has very thin leads, looks more like 1N4148’s leads.
Without diode it’s better, module responds at+ccid, indication SIM is wired correctly, at+cpin=”xxxx” correctly with OK and doesn’t restart, momentarily takes a lot of juice from power supply (that is expected) though any consequent operation (+cpin? +cops? +ccid) above SIM card ends up with ERROR message until I restart the module manually. Meh. Perhaps it can’t access network, it may be because of the “antenna”, which is in the same category as the power supply.

EDIT: got it working by using longer piece of wire antenna instead of original short piece of wire. One has to use proper chinese engineering when dealing with chinese devices. Sometimes I get garbled characters, though http://pastebin.com/tRxLm7YV
I think those modules could be usable with proper power supply and antenna, but that goes for all GSM modules. I like Antenova A10340 antennas https://picasaweb.google.com/1001130002 … tK_Wi6D-fw

jaromir wrote:The first one, with crusty and overheated module is dead as brick.

While g..ling the M590E module I’ve seen a few schematics there. A good decoupling on any pin seems to be a must with this kind of modules. For example a design suggests 33pF-100pF on any pin (serial i/o too). Also a table in an another one shows 3 capacitors wired on vbat – 100pF, 100nF and C1, where C1=10uF for 2A tx peaks … 1000uF for 0.5A peaks (! interesting).
Such combination is btw. always recommended by big players as decoupling with ie. 1000uF aluminum works maybe till 200Hz cutoff.. It will not provide enough charge with milli/useconds current pulses.. A ceramic multilayer of 10uF capacitance is a better choice (for high current, short rising edge current pulses) than the 1000uF one, sure ..

jaromir wrote:One has to use proper chinese engineering when dealing with chinese devices.

Pito wrote:
Such combination is btw. always recommended by big players as decoupling with ie. 1000uF aluminum works maybe till 200Hz cutoff.. It will not provide enough charge with milli/useconds current pulses.. A ceramic multilayer of 10uF capacitance is a better choice (for high current, short rising edge current pulses) than the 1000uF one, sure ..

Sound advice jaromir good decoupling is vital where there is RF involved. Screening the RF away from other components is also a good idea, but in this case, that has been taken care of by the module. There’s a reason why its in that tin can, it keeps the pixies running up and down the antenna, rather than having them scurry all over the show.

I once spent a couple of months trying to figure out why one of my customer’s very expensive computers would crash randomly especially in the evening and early morning. The problem, it turned out was due to the Taxi firm operating a very large VHF transmitter in the adjacent building. We moved the computer behind a large earthed metal filing cabinet and the issue went away. History does not relate whether there were any long term health effects from being gently cooked by close proximity to the transmitter, but it certainly didn’t do the computer any good. Random crashes due to RF bleeding into your microntroller are something you might have to contend with if you don’t keep on top of the screening and decoupling.

I will try to clean up both my modules and try them later today.

I have solder wick etc, so it should be ok.

I just object to being miss sold.

Re:nokia displays

Yes…
Same thing.

In have received many that are scratched, or dirty, or even have burn in with an fuzzy logo.

Only half of them could be considered as vague new and possibly used in a project that I build for someone else.

ahull wrote: The problem, it turned out was due to the Taxi firm operating a very large VHF transmitter in the adjacent building. We moved the computer behind a large earthed metal filing cabinet and the issue went away.

ahull wrote: The problem, it turned out was due to the Taxi firm operating a very large VHF transmitter in the adjacent building. We moved the computer behind a large earthed metal filing cabinet and the issue went away.

Long time back one of my clients had the similar issue – at mornings a segment of their corporate network was shut down.
After a significant troubleshooting effort they discovered a cleaning lady pulled a network component’s mains plug off while hoovering the carpet

BTW, does the stuff work with those cheapo Tesco cards? (Cheapo.. not at all .. a pity the credit is valid for 30days only and it is not transferrable )..

pito: don’t know about where are you located, but here in slovakia O2 offers some good monthly rates.
I set-up my second “testing” SIM card from O2 in 2014 or so, made calls/sent SMS for something like 1EUR in total while testing my cellphones, now I turned it on after a year of inactivity, the number is still alive. I didn’t pay a single cent in the period of inactivity. Now I made a few calls/sent a few SMS, I expect invoice for 10 cents perhaps.

Interesting, indeed. Currently on Rapa Nui, but I’ll travel to EURO next week, so maybe I’ll give a shot to O2 there..
BTW, what could be done with the M590E module (I mean what kind of services/connectivity may work):
1. when I get a sim with voice only
2. when I get a sim with voice and “data” ?

Pito wrote:Long time back one of my clients had the similar issue – at mornings a segment of their corporate network was shut down.

Pito wrote:…Currently on Rapa Nui…

Same Pito from Arduino.cc

pito.jpg (2.1 KiB) Viewed 663 times

Computer tape drive. With tape loops in vacuum columns. Mounted in military electronics shelter near door.
Every morning – no workie.
Every late morning – workie fine.

Eventually, I realized: Early AM, sun shined in, hit tape loop reflector sensors. Screwed up vacuum control.
Cure: Cardboard taped to drive’s plexiglass door.

I have more.

stevech wrote:Computer tape drive. With tape loops in vacuum columns. Mounted in military electronics shelter near door.
Every morning – no workie.
Every late morning – workie fine.

Eventually, I realized: Early AM, sun shined in, hit tape loop reflector sensors. Screwed up vacuum control.
Cure: Cardboard taped to drive’s plexiglass door.

I have more.

ahull wrote:
Keep them coming… this thread is already way off topic

Kenjutsu wrote:Any links to the AI Tinker A6 module documentation, and / or pinout of the breakout board?

FYI –

http://shortn0tes.blogspot.cz/2016/05/n … g-and.html

Pito wrote:FYI –

http://shortn0tes.blogspot.cz/2016/05/n … g-and.html

Umm

Good idea about looking at photos of the assembled board.

I must admit, I stalled when I realised that the board doesnt have any markings to indicate capacitor polarisation.

I’d already soldered on the SM cap before I realised that it was polarised.

Luckily I think its the right way around.

RogerClark wrote:Luckily I think its the right way around.

Thanks Andy

The cap looks the same way around as I’ve seen it in other photos.

I will try to build my board at the weekend.

Well,

Good job I bought 2, as the SM resistors are so small, that one if them got catapulted out of its tap, and I’ve no idea where it went (looked high and low).

Another tip. As far as I can tell the LED has its positive side connected to the GSM module, so I used my multimeter in resistance mode to work out which way around was which on the little SMD LED, as I couldn’t see any markings on it, even using a 10x magnifier.

I’ve not powered it up yet as I need to find a PSU that will supply enough current

RogerClark wrote:
<…>
that one if them got catapulted out of its tap, and I’ve no idea where it went (looked high and low).
<…>

RogerClark wrote:I couldn’t see any markings on it, even using a 10x magnifier

So, another three pieces of M590 modules from different seller arrived.
Again, all three modules are desoldered, though not as butchered as the first one. This time even the 100uF tantalum caps are recycled

So, I have a few of those modules, probably will make proper dedicated board for that.

I forgot to post, that only 1 of the 2 modules I bought, contained all the components.

One of them did not contain the Sim card holder, so it would have been useless.

However the vendor didnt bother to respond about this, or to my negative review.

I received yesterday two M590 diy kits from the same seller, the modems with different revision numbers, refurbished (ugly soldering rests on the pins).
Nevertheless, they both seem to work (at least the serial communication is alive).
However, my problem is that I have very week GSM reception in my testing room.

Do you know any possibility how to improve GSM reception without leaving the room?
Antenna? What kind of? Any recommendation/experience?

stevestrong wrote:
Do you know any possibility how to improve GSM reception without leaving the room?
Antenna? What kind of? Any recommendation/experience?

stevestrong wrote:
<…>
Do you know any possibility how to improve GSM reception without leaving the room?
Antenna? What kind of? Any recommendation/experience?

So, another three pieces of M590 modules from different seller arrived.
Your massive acquisitions could increase the price of the stuff, please stop it!!!

How weak:

+CSQ:99,99

Would it make sense to place the delivered small antenna close to the window and use a 3m coax cable to connect it to the module?

In summary, for the tl;dr brigade, try a length of stiff copper wire just over a foot long, and snip of small pieces till you figure out the best length.

Now for the rest of you with working brains and longer attention spans.

I had a brief look at this problem a few days back, and came up with a couple of possibilities.

The “random long wire” antenna would probably boost reception fairly easily, but would need careful tuning to boost transmission.

There are a couple of problems with this approach, particularly if you need to get your device certified, This is not an issue for DIY projects, but ALL mobile phone/gprs devices are *meant* to adhere to the specs for GPRS radiated power etc, which are at least in part, designed to avoid you frying your brains if you strap the thing next to your head. This is not a very likely scenario, give the amount of power involved, but if you build a highly directional antenna, and work with higher power levels, the local radiated power can be high enough to be a worry. RF burns are not pleasant, particularly if you manage to induce them in your body tissues.

You can purchase antennae online through from the usual suspects, or you can design your own.

Here are a couple of links that might prove useful for constructing a DIY antenna.

http://electronics.stackexchange.com/qu … cb-antenna

https://en.wikipedia.org/wiki/Numerical … etics_Code

Although RF design can be a bit of a black art, the basic idea is to construct a radiating element or multiple elements that has/have the correct physical dimensions to match the wavelength or an exact binary fraction (1/2, 1/4, 1/8 etc) or multiple (1,2,4,8 etc) of the wavelength you require. This is because we want the element(s) to resonate (think of this as being the radio equivalent of a tuned metal bar struck by a hammer producing a good loud ringing noise at a single pitch). If you have multiple elements, they all have to resonate together, and this is where things start to get complicated.

http://www.csgnetwork.com/freqwavelengthcalc.html

Using the above calculator and assuming a frequency of 900 MHz, we can see that a 1/4 wave antenna element would be around 83mm or 3 /14″ so a simple single element antenna would consist of a piece of wire carefully snipped to slightly over the correct length, then “trimmed” by snipping or filing off small amounts to achieve the best results. Since we don’t have a SWR meter or radiated power measuring capabilities, we would need to judge the results based on the signal strength reported by the module, but that should be “good enough for government work” as they say.

The reason I suggested a piece of wire just over a foot long at the start should now be clear, a single wavelength at 900 MHz is around 13 inches, or just over a foot long. A better bet might be to take a 16″ length of wire, and form a few turns at the bottom, then stretch and shrink the length by shorting out these turns with another bit of wire. Crude but effective. Placing a suitable value trimmer capacitor across the turns and/or adding a movable ferrite core to them would be even better, allowing trimming without physically removing any of the copper.

The larger the antenna, the better it performs (within certain limits, and assuming it adheres to the rule about physical dimensions approximating fractions or multiples of wavelength), so a 1/2 wave, or full wavelength element will typically work better than a 1/4 or 1/8 etc. Big bells resonate better than small ones.

Back in my youth, I used to build CB (27MHz) antennae which had a typical radiating element of around 2m, although the actual wavelength at 27 MHz is nearer 11m, so they required a “loading coil”, and tuning to make the whole thing behave as if it were 11m long.

With care, and because someone bet me I couldn’t. I managed to achieve a good SWR from a hand constructed coil and a radiating element that was a carefully tuned drawing pin. This did get pretty hot, and was probably not doing the RF stage of my transmitter any favours, but no magic smoke was released, and the thing could transmit over reasonable distances.

If you touched it, or even breathed on it, the SWR fluctuated wildly, which comes to my second point, the antennae should be free from contact with other objects that have any measurable impedence (which includes you), and a good ground plane, or reflector can make a huge difference to the SWR and the amount of radiated power that actually leaves the antenna, and to the amount of signal that it receives. If you, or anything with other than perfect insulating properties, touches the radiating element (or indeed gets sufficiently close to it, due to capacitive/inductive coupling and the resultant reactance), you will de-tune it, and it will stop resonating at the desired frequency (think of the ringing metal bar, or a musical instrument string, if you touch them while they are resonating, you badly affect the resulting tone). This is one reason why most commercial GPRS antennae are encapsulated in plastic or silicone.

DIY antenna for this particular purpose is not easy to make. CB @27MHz is not 900MHz or something like that. You definitely would need a good equipment and fair understanding what you actually doing. To tune a piece of wire to 900MHz at a random quality coax, such it matches 50ohm PA, is a matter of a pure luck.
I would suggest to purchase a ready made patch antenna or a Ground Plain whip, connected via 20-30cm of quality 50ohm coax (usually a part of the antenna).
There are dozens of GSM antennas (with coax and connector, 1-8dBi gain) available for a few $.

Pito wrote:DIY antenna for this particular purpose is not easy to make. CB @27MHz is not 900MHz or something like that. You definitely would need a good equipment and fair understanding what you actually doing. To tune a piece of wire to 900MHz at a random quality coax, such it matches 50ohm PA, is a matter of a pure luck.
I would suggest to purchase a ready made patch antenna, connected via 20-30cm of quality 50ohm coax (usually a part of the antenna).
There are dozens of GSM antennas (with coax and connector, 1-5dBi gain) available for a few $.

Since you can never have too many pieces of test equipment, and since there seems to be some interest in building antennae, you might like to take a look at … http://fpvlab.com/forums/showthread.php … lf-project

The secret sauce if you want to build a SWR meter for higher frequencies is a microwave detector diode. Something like this…

http://www.ebay.co.uk/itm/Avago-Agilent … dl6lUYmpnw

… and of course the ability to solder very small things, but I think if you managed to assemble the “GSM GPRS Module at 1.93$!!!!!” you are already over qualified.

If you do decide to build rather than buy an antenna, then you can make it highly directional, and as large as you please (visits from the local regulatory authorities notwithstanding of course ), which may be the only practical way you can get a usable signal in a remote location.

Do bear in mind, I am not encouraging you to flout the appropriate regulations in your country, so obviously you proceed at your own risk.

If you don’t want to build, as Pito pointed out, there is always ebay to fall back on.

However I cannot vouch for the claims of any of these devices, so try them and see if they live up to their spec. Look for low VSWR (<1.5) and high gain (5dB or better), and the correct frequency bands (900/1800MHz, *NOT* 433MHz) not that you will be able to verify these claims without a SWR meter of course

Perhaps something like this… http://www.ebay.co.uk/itm/Magnetic-Base … Sw~OVWy1oL .. would be a good starting point.

I would go with those green pcb antennas – like this one:
http://www.ebay.co.uk/itm/2G-3G-3DBI-PC … Swv-NWYTX-
and a lot of others, with max 10-30cm long coax.
Why? 0.1inch plays a significant role when you tune an UHF antenna (you may go fully off, see ie. http://colinkarpfinger.com/blog/2010/th … na-design/ even the guy tunes 2.4GHz one), so it seems to me a pcb etched antenna is more precise than a whip antenna with loading coil.. 1-3dBi is enough (still much better than a bended paper clip).
3m long RG174 coax or, RG58 one – try to avoid it for any price

Thanks for many precious suggestions, I’ve chosen the easiest solution: http://www.aliexpress.com/item/Short-GS … 78218.html
I will cut the 3m coax if the attenuation will be too high, and try to place the module close to the window, although this move will seriously decrease the WAF …

stevestrong wrote:How weak:

+CSQ:99,99

Would it make sense to place the delivered small antenna close to the window and use a 3m coax cable to connect it to the module?

simonf wrote:stevestrong wrote:How weak:

+CSQ:99,99

Would it make sense to place the delivered small antenna close to the window and use a 3m coax cable to connect it to the module?

GSM bands from wikipedia:
GSM bands.JPG

The dual band, multiple frequency nature of GSM does mean that any design will be a compromise, and you can see that in most cases the designers of PCB based designs have chosen to make their design effectively two antennae in one.

PS: what about using a ceramic multilayer GSM antenna from a scrap mobile phone (wired via 10-20cm long coax)?

This would work, and has the advantage of being a tried and tested design, however it will obviously be no better in terms of performance than the original mobile phone. Some of the larger designs, or indeed a simple full wave or half wave dipole centred roughly on the 900 MHz centre frequency may in fact perform significantly better. In other words two very long paper clips may actually be as good as the ceramic multilayer device, and two 13″ straight wires arranged as dipole may be better still. If you need to use a directional antenna in order to cover larger distances, you might have to try a yagi or some other variant, which would give you the ability to point more of the available RF energy at your distant cell tower. This however leads to greater complexity, and potentially may fall foul with the authorities if you are not careful.